Re: A sketch of Old Albic 1/4: Phonology

Hallo!

On Mon, 21 Jun 2004 16:33:36 EDT,
David Peterson <ThatBlueCat@...> wrote:

> Joerg wrote:
>
> <<Most are as in IPA, except: c = /k/, ph = /f/, th = /T/, ch = /x/,
> ñ = /N/, v = /w/.>>
>
> Just to be clear, the letter enya is used for angma, that is the "ng" in
> "ring"?
Yep.

> <<The stops (p, t, c, b, d, g) have fortis and lenis allophones.
> The lenis allophones occur between vowels, semivowels and liquids,
> the fortis allophones in all other environments.>>
>
> What does this mean phonetically?   Is there a single definition of lenis vs.
> fortis?
What this means is basically that the fortis stops are pronounced
with somewhat more force than the lenis ones.  Later, the lenis
voiced stops became voiced fricatives, and the lenis voiceless
stops either voiced stops or voiceless fricatives, depending on
the dialect in question.  What you can see here is the first step
towards a system of Celtic-like initial mutations.  Consider, for
example, the noun phrase _a catha crara_ `the red cat', in which
the initial consonants of both _catha_ and _crara_ are lenis
because they follow a vowel in a linked word.  So when the lenis
stops are voiced, the result is _a gatha grara_, with initial
mutations of both _catha_ and _crara_.

> <<Some affixes undergo vowel harmony: the vowel in the affix always
> matches the nearest vowel of the stem.>>
>
> This might be better explained as reduplication.
It is surely different from the vowel harmony systems found in
Uralic or Altaic languages, where the vowels in the affixes are
underspecified only with regard to one or two features, but
I think "vowel harmony" fits the phenomenon better than
"reduplication" because it is indeed an underspecified vowel
in the affix, and not entire morae, syllables etc. that are
repeated.

> <<In words with three or more syllables, the accent
> falls on the antepenultimate (third-last) syllable if both the
> penultimate and ultimate syllable are light (i.e., they are open and
> have a short vowel), otherwise on the penultimate syllable.>>
>
> So a word like /panatan/ would be [pa.'na.tan]?
Yes.  This is actually the one case where the Old Albic accent rule
differs in its outcome from the Latin (and Quenya) rule.

> <<This accent rule can be formulated in a more concise manner using the
> concept of the mora. A mora is a metric unit below the syllable. A
> light syllable consists of one mora, a heavy syllable of two. In light
> of this, it is the third-last mora that carries the accent in Old
> Albic.>>
>
> This strikes me as odd.   Any natlangs that count morae in this way?
I have heard about the existence of mora-counting languages, though
I am no expert on that.  I have seen a description of the Latin
accent system in terms of morae once, it amounted to saying that
the accent falls on the antepenultimate mora if one assumes that
final syllables always have only one mora even if they are heavy.
The latter is clearly an artificialism.  But I see no reason
why a language should not have an accent that always falls on the
antepenultimate mora.

> Neat stuff!   On to your other posts...
Thanks!

Greetings,

Jörg.